Expression and role of the long noncoding RNA Linc00909 in chemoradiotherapy for colorectal cancer

Abstract

BACKGROUND

Colorectal cancer is one of the major malignant tumors worldwide. Previous studies have shown that long noncoding RNAs have significant effects on various human cancers, but the underlying molecular mechanisms remain unclear.

AIM

To investigate the impact of the long noncoding RNA Linc00909 on colorectal cancer and its molecular mechanism.

METHODS

Using 196 paired colorectal cancer and adjacent tissues from 30 patients with rectal cancer (19 resistant, 11 sensitive) to neoadjuvant chemoradiotherapy, we assessed Linc00909 expression using real-time quantitative polymerase chain reaction and correlated it with survival outcomes. Functional assays in Linc00909-overexpressing SW620 and DLD1 cells—including Cell Counting Kit-8 assays, flow cytometry, and sphere formation—evaluated proliferation, cell cycle, apoptosis, stemness, and 5-fluorouracil/radiotherapy sensitivity. Tumorigenicity was examined in vivo. Bioinformatic and molecular experiments explored Linc00909’s competing endogenous RNA mechanism.

RESULTS

Linc00909 was upregulated in resistant and tumor tissues and independently predicted poorer overall and disease-free survival. Linc00909 overexpression enhanced proliferation, colony formation, antiapoptotic capacity, stemness, and therapy resistance in vitro and promoted tumor growth in vivo. miR-625-5p was bioinformatically predicted and experimentally validated as a binding target of Linc00909, with its expression decreasing upon Linc00909 overexpression. SOX2 was identified as a downstream target of miR-625-5p and showed increased expression at both the messenger RNA and protein levels following Linc00909 upregulation.

CONCLUSION

Linc00909 can affect the proliferation, colony formation, apoptosis, cell cycle regulation, spheroid-forming ability, and sensitivity of colorectal cancer cells to 5-fluorouracil and radiotherapy. Mechanistically, Linc00909 may regulate the expression of the downstream target gene SOX2 by sequestering miR-625-5p, thereby influencing the sensitivity of colorectal cancer cells to radiotherapy and chemotherapy. Therefore, Linc00909 can be considered a potential therapeutic target for treating colorectal cancer.

Key Words: Colorectal cancer; Linc00909; Neoadjuvant chemoradiotherapy; Prognosis; Therapy resistance

Core Tip: Linc00909 is elevated in therapy-resistant and tumor tissues of colorectal cancer and serves as an independent indicator of inferior overall and disease-free survival. Its overexpression promotes proliferation, clonogenicity, apoptosis evasion, stem-like traits, and resistance to 5-fluorouracil and radiotherapy in vitro, while accelerating tumor growth in vivo. Mechanistically, Linc00909 sponges miR-625-5p, thereby relieving repression of SOX2 and activating a Linc00909/miR-625-5p/SOX2 axis that drives therapeutic resistance. This axis is a candidate target for improving colorectal cancer treatment.

INTRODUCTION

Colorectal cancer is one of the most common types of malignant tumors, with its incidence and mortality ranking third among all malignant tumors[1]. In China, its incidence has been increasing annually over the past 10 years, seriously affecting people’s health and quality of life. According to the 2020 edition of China’s “Diagnosis and Treatment Specifications for Colorectal Cancer”[2] and the 2020 National Comprehensive Cancer Network guidelines for the diagnosis and treatment of rectal cancer, patients with locally advanced middle and low rectal cancer should undergo neoadjuvant radiotherapy and chemotherapy before surgery, followed by radical surgical treatment. Studies have shown that adopting this treatment plan has obvious advantages in terms of tumor shrinkage, tumor downstaging, improvement in the local control rate, improvement in the sphincter preservation rate, and reduction in local recurrence in patients with colorectal cancer[3-5]. However, owing to individual differences, patients’ sensitivity to radiotherapy and chemotherapy varies, and some patients experience treatment failure and disease recurrence due to resistance to radiotherapy and chemotherapy, greatly affecting the cure rate. Therefore, for patients who are resistant to radiotherapy and chemotherapy, the benefits are limited, but they must bear the adverse consequences of treatment side effects and disease progression[6]. Studies have shown that the 10-year disease-free survival (DFS) rate of patients with a complete response after radiotherapy and chemotherapy is 89.5%, whereas that of patients with resistance to radiotherapy and chemotherapy is only 63%[7]. Moreover, the pathological response status of patients determines subsequent treatment plans and thus affects their prognosis and quality of life[8]. Therefore, exploring the mechanisms underlying colorectal cancer sensitivity to radiotherapy and chemotherapy and identifying new therapeutic targets to overcome resistance and increase sensitivity may enable patients to obtain greater benefits.

Long noncoding RNAs (lncRNAs) refer to a class of long-chain RNA molecules with a transcript length greater than 200 nucleotides that do not encode proteins. Most are transcribed by RNA polymerase II and undergo posttranscriptional modifications. They can be located in the nucleus or cytoplasm. Their types and structures are diverse. They regulate the expression of related genes at multiple levels, such as the pretranscriptional, transcription, transcriptional, and posttranscriptional levels, and are widely involved in the pathological and physiological processes of the body, possessing extremely important and complex biological functions[9-11]. In recent years, with the widespread application of next-generation sequencing technology, an increasing number of studies have shown that lncRNAs are closely related to the occurrence and development of colorectal cancer as well as sensitivity to radiotherapy and chemotherapy[12,13]. Current studies suggest that lncRNAs can interact with proteins, messenger RNAs (mRNAs), and microRNAs (miRNAs) in the form of competing endogenous RNAs (ceRNAs) and “miRNA sponges” and can also act as major transcription factors to participate in the occurrence, development, drug resistance, and metastasis of colorectal cancer[14]. Studies have shown that lncRNAs are also involved in the regulation of radiotherapy and chemotherapy in rectal cancer and can participate in the complex regulatory network of radiotherapy and chemotherapy sensitivity in rectal cancer through pathways involving apoptosis, drug metabolism, and epithelial-mesenchymal transition, thereby affecting the sensitivity of rectal cancer to radiotherapy and chemotherapy[15-17]. Some studies have reported that lncRNAs are related to the sensitivity of rectal cancer to radiotherapy and chemotherapy[18-21]. These dysregulated lncRNAs may participate in the regulation of rectal cancer cell sensitivity to radiotherapy and chemotherapy through one or more of the above mechanisms. However, the specific identities and roles of lncRNAs in these regulatory networks have not yet been fully elucidated, and the complete mechanisms remain to be discovered. Therefore, further research is needed to address this problem. Wang et al[22] reported that the lncRNA HNF1A-AS1 is not only highly expressed in non-small cell lung cancer tissues and cells but can also regulate the expression of downstream mitogen-activated protein kinase kinase 4 (MAP2K4) by sequestering miR-92a-3p.

MAP2K4 can affect the radiotherapy sensitivity of non-small cell lung cancer cells by regulating the MAP2K4/c-Jun N-terminal kinase (JNK) pathway. Li et al[23] reported that the lncRNA LHFPL3-AS1 is highly expressed in oral squamous cell carcinoma tissues and cell lines. Moreover, its expression is upregulated in cisplatin-resistant oral squamous cell carcinoma cell lines. Patients with high expression of the lncRNA LHFPL3-AS1 have a worse prognosis. The lncRNA LHFPL3-AS1/miR-362-5p/CHSY1 signaling pathway plays an important role in regulating the development of oral squamous cell carcinoma and cisplatin resistance. Silencing of the lncRNA LHFPL3-AS1 can not only inhibit the proliferation, migration, and invasion of oral squamous cell carcinoma cells but also reduce their cisplatin resistance. These studies suggest that lncRNAs are involved in the complex regulatory network of tumor sensitivity to radiotherapy and chemotherapy, but the roles and specific mechanisms of lncRNAs in the sensitivity of rectal cancer to radiotherapy and chemotherapy are not yet thoroughly or systematically understood, and further research is needed. Therefore, further identification of new therapeutic targets to overcome radiotherapy and chemotherapy resistance and increase patient sensitivity to these treatments may lead to greater benefits.

By analyzing gene chip data from 31 patients with rectal cancer obtained by our research group in a previous study, we identified several lncRNAs that were significantly differentially expressed between the neoadjuvant chemoradiotherapy (NCRT)-resistant group and the NCRT-sensitive group. Linc00909 was one of these lncRNAs. We subsequently searched for related reports and found that there are currently few reports on Linc00909. It has been reported only in glioma and colorectal cancer. In human glioma, Linc00909 functions as an oncogene. Fluorescence in situ hybridization results suggest that Linc00909 is located in the cytoplasm and can act as a “miRNA sponge,” competing with miR-194 to cause the upregulation of the downstream target molecule mucin 1 C-terminal subunit (MUC1-C) and promoting the progression of glioma[24]. Xu et al[25] reported that high serum Linc00909 expression can be an effective diagnostic biomarker for colorectal cancer; however, its mechanism of action has not been further verified through in vivo and in vitro experiments. Therefore, to understand the role of Linc00909 in colorectal cancer, we explored it through the following experimental methods. First, we used real-time quantitative polymerase chain reaction (RT-qPCR) to detect differences in Linc00909 expression between NCRT-resistant and NCRT-sensitive tissues, as well as differences in expression between cancer tissues and adjacent tissues, and to evaluate the role of Linc00909 in predicting sensitivity to radiotherapy and chemotherapy and patient prognosis. In vitro cell experiments and subcutaneous tumor formation experiments in nude mice clarified the biological functions of Linc00909 in the occurrence and development of colorectal cancer and the regulation of sensitivity to radiotherapy and chemotherapy. Finally, bioinformatics analysis preliminarily explored the possible molecular mechanisms of Linc00909 in the occurrence and development of colorectal cancer and the regulation of sensitivity to radiotherapy and chemotherapy. The results of these experimental studies suggest that to some extent, Linc00909 can affect the occurrence and development of colorectal cancer and sensitivity to radiotherapy and chemotherapy and provide experimental evidence that Linc00909 may be useful for predicting sensitivity to radiotherapy and chemotherapy and evaluating patient prognosis in clinical practice.

MATERIALS AND METHODS

Sample collection

Patients diagnosed with colorectal cancer at the First Affiliated Hospital of Fujian Medical University were selected on the basis of pathological and imaging findings from January 2017 to December 2017. All patients received neoadjuvant radiotherapy and chemotherapy [with a total dose of 45.0-50.4 Gy for the primary tumor and high-risk areas, with each fraction ranging from 1.8 to 2.0 Gy, for a total of 25-28 fractions; during radiotherapy, 5-fluorouracil (5-FU) or capecitabine monotherapy was given concurrently]. Thirty patients meeting the criteria were included. The tumor response after radiotherapy and chemotherapy was evaluated according to the tumor regression grade. The specific criteria were as follows: Grade 0, complete regression with no residual tumor cells; grade 1, moderate regression with single or small foci of residual tumor cells; grade 2, slight regression with residual tumor tissue and a large amount of fibrotic stroma; and grade 3, no regression with extensive residual tumor tissue and no or little tumor cell necrosis. Among them, 11 patients had tumor regression grades of 0 or 1 and were included in the NCRT-sensitive group; 19 patients had tumor regression grades of 2 or 3 and were included in the NCRT-resistant group.

Additionally, 196 pairs of colorectal cancer tissue and adjacent normal tissue samples were collected from surgeries performed at the First Affiliated Hospital of Fujian Medical University from January 2017 to December 2017, and the corresponding clinicopathological information and follow-up data were also collected. The follow-up period for patients after surgery ended on October 31, 2020. Overall survival (OS) was defined as the time from surgery to the end of follow-up or death, measured in months. DFS was defined as the time from the date of surgery to disease recurrence or death due to disease progression, measured in months. All patients provided written informed consent (approval No. MTCA, ECFAH of FMU [2015]084-2).

Total RNA extraction and RT-qPCR

In accordance with the TRIzol reagent instructions (Invitrogen, Carlsbad, CA, United States), total RNA was extracted from SW620 and DLD1 cells. Complementary DNA was synthesized using PrimeScript™ II Reverse Transcriptase (TaKaRa, Shiga, Japan). RT-qPCR was performed on an ABI Prism 7900HT sequence detector (Applied Biosystems, Foster City, CA, United States) using a SYBR Green I kit (TaKaRa), and relative expression was calculated using the 2^−ΔΔCt method. In the relevant assays, glyceraldehyde-3-phosphate dehydrogenase and U6 were used as internal controls.

Cell culture

Two human colorectal cancer cell lines (SW620 and DLD1) were purchased from the American Type Culture Collection in Manassas, United States, and cultured in Dulbecco’s modified Eagle’s medium (Corning, Tewksbury, MA, United States) supplemented with 10% fetal bovine serum (Gibco, Gaithersburg, MD, United States) and 1% penicillin-streptomycin (Gibco). The cells were cultured at 37 °C in 5% carbon dioxide, with the medium replaced every 2-3 days. All cell lines used in this study were authenticated using short tandem repeat profiling by Wuhan Zixian Biotechnology Co., Ltd. (Hubei Province, China) before experimentation. The cell lines were regularly tested for mycoplasma contamination and confirmed to be negative.

Overexpression of LINC00909 with a lentivirus

The LINC00909 sequences used were as follows: Control, LINC00909; F, CTTTTTTGTTAGACGGATCCGAAGGACTTCCGGTGGCTTCCAAGG; and R, AAAGATATTTTATTACCGGTTTAC. Lentiviral virions were produced by cotransfecting HEK293T cells with 5 μg of the PLVX-puro vector and 5 μg of the packaging and envelope vectors using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s protocol. The lentiviral virions were then harvested 48 hours after transfection. SW620 and DLD1 cells were separately infected with lentiviruses containing ov-LINC00909 or ov-control for 24 hours. Two days later, the virus-infected cells were selected with 4 μg/mL puromycin (Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) for 48 hours and subjected to subsequent assays.

Colony formation assay

Cells in the logarithmic growth phase were trypsinized, resuspended, and seeded into 6-cm dishes at 1000 cells/dish. After 2-3 weeks of culture, the colonies were fixed with 4% paraformaldehyde, stained with crystal violet, and counted. The colony formation rate was calculated as (number of colonies/number of seeded cells) × 100%.

Resistance to 5-FU

Anchorage-dependent cell growth was evaluated using a Cell Counting Kit-8 (CCK-8) (Takeda Laboratory Co., Ltd., Osaka, Japan) according to the manufacturer’s instructions. The cells were seeded at a density of 3 × 10³ cells per well in a 96-well plate. When the cells reached 60% confluence, the culture medium was removed and replaced with fresh culture medium containing different concentrations of 5-FU. The cells were subsequently cultured for another 48 hours. Absorbance was measured at 450 nm using a microplate reader, and cell viability was subsequently calculated.

Cell proliferation assessment via the CCK-8 assay

Cells in the logarithmic growth phase were trypsinized, counted, and seeded into 96-well plates at a density of 3000 cells per well (n = 6). After incubation for specified durations (4 hours, 24 hours, 48 hours, 72 hours, 96 hours, and 120 hours), 10 μL of CCK-8 reagent was added to each well, and the plates were further incubated for 2 hours. Absorbance was measured at 450 nm using a microplate reader. The cell proliferation rate was calculated as (A_d − A_blank_d)/(A_4h − A_blank_4h). All experiments were performed in triplicate.

Flow cytometry assay

Apoptosis of SW620 and DLD1 cells was monitored using an Annexin V-FITC/PI staining kit (Invitrogen) according to the manufacturer’s instructions. In total, 1 × 10⁶ cells were seeded in 6-well plates and double-stained for 15 minutes in the dark. The results were examined using a flow cytometer (BD Biosciences, Franklin Lakes, NJ, United States).

Sphere formation assay

The cells were collected, washed with phosphate-buffered saline, and then resuspended in stem cell culture medium (Dulbecco’s modified Eagle’s medium/F12 supplemented with epidermal growth factor, basic fibroblast growth factor, N2 supplement, and B-27) at a density of 1 × 10⁴ cells/mL. In total, 1000 cells were seeded per well into ultralow-attachment 24-well plates in triplicate. After uniform distribution, the cells were cultured at 37 °C in 5% carbon dioxide. Half of the medium was replaced daily. Sphere formation was monitored, and the diameter of the spheres was measured under a microscope after 1-2 weeks of culture.

Western blot

The total protein content in the tissues and cells was quantified, followed by protein transfer onto a polyvinylidene fluoride membrane. The membrane was blocked at room temperature for 2 hours and then incubated with primary antibodies against SRY-box transcription factor 2 (SOX2) (1:3000) and tubulin (1:5000) at 4 °C overnight. After washing, the membrane was incubated with the corresponding secondary antibody for 2 hours. The protein bands were visualized using a chemiluminescence substrate and quantified with ImageJ software (National Institutes of Health, Bethesda, MD, United States).

Animal experiments

SW620 and DLD1 cells stably overexpressing Linc00909 or empty vector controls were harvested during the logarithmic growth phase, resuspended in phosphate-buffered saline at a concentration of 1 × 10⁸ cells/mL, and kept on ice before injection. In total, 20 male BALB/c nude mice were randomly divided into 2 groups (n = 5 per group): The vector control group and the Linc00909 overexpression group. Each mouse was subcutaneously injected with 100 μL of cell suspension (1 × 10⁷ cells) in the axillary region. Tumor growth was monitored every 3 days, and tumor volume was calculated using the following formula: Volume = 0.5 × L × W². After 5 weeks, all mice were euthanized by cervical dislocation. The tumors were excised, weighed, and processed for further analysis by either freezing at −80 °C or fixation in formalin (approval IACUC FJMU No. 2023-Y-0706).

Online databases and web resources

Potential miRNA binding sites for Linc00909 were predicted using the miRcode database. The target genes of these miRNAs were subsequently identified using MiRTarBase, TargetScan (version 7.1), and miRDB. The overlapping predictions from these 3 databases were considered high-confidence targets. A ceRNA network was visualized using Cytoscape. The following online resources were used in this study: MiRcode (http://www.mircode.org), MiRTarBase (http://mirtarbase.mbc.nctu.edu.tw), TargetScan (http://www.targetscan.org), and miRDB (http://www.mirdb.org).

Statistical analysis

Statistical analysis was performed using GraphPad Prism 7.0 or SPSS 23.0. The data are presented as mean ± SD. Differences between groups were assessed using Student’s t-test or analysis of variance. P < 0.05 was considered statistically significant.

RESULTS

Expression of Linc00909 in rectal cancer tissues resistant to chemotherapy and radiotherapy

In the previous stage of this research project, our team analyzed 31 rectal cancer biopsy samples from the colorectal surgery department of our hospital before radiotherapy and chemotherapy. Using Affymetrix gene chips, we detected key responsive genes and their related pathways. By analyzing the gene expression profiles, we identified a total of 18419 genes, including 241 lncRNAs. Cluster analysis of the lncRNA expression profiles revealed a subset of genes whose expression significantly differed between the chemosensitive and chemoresistant groups, among which Linc00909 was selected for further investigation (Figure 1A). To further validate the potential of Linc00909 in predicting chemoradiation sensitivity, we analyzed 30 rectal cancer samples from this cohort, including 19 from nonresponders and 11 from responders to NCRT. The RT-qPCR results demonstrated that Linc00909 expression was significantly upregulated in NCRT-resistant tissues (ΔCt = 7.59 ± 0.37) compared with NCRT-sensitive tissues (ΔCt = 10.02 ± 0.73, P < 0.01) (Figure 1B). Moreover, Linc00909 expression was significantly upregulated in tumor tissues (ΔCt = 8.78 ± 0.16) compared with adjacent normal tissues (ΔCt = 10.64 ± 0.12, P < 0.001) among 196 paired colorectal cancer samples, as determined by RT-qPCR (Figure 1C). In addition, receiver operating characteristic curve analysis revealed that Linc00909 effectively distinguished between NCRT-resistant and NCRT-sensitive patients, with an area under the curve of 0.73 (Figure 1D). The association between Linc00909 expression and patient survival was analyzed using log-rank tests based on follow-up data, and Kaplan-Meier curves were generated (Figure 1E and F). High Linc00909 expression was significantly correlated with shorter OS and DFS (P < 0.01).

Figure 1 Linc00909 expression in therapy-resistant rectal cancer tissues. A: Volcano plot of long noncoding RNA expression; B: Differential expression of Linc00909 in neoadjuvant chemoradiotherapy-resistant vs neoadjuvant chemoradiotherapy-sensitive tissues; C: Differential expression of Linc00909 in rectal cancer samples; D: Receiver operating characteristic analysis of Linc00909 in predicting chemoradiotherapy sensitivity; E and F: Kaplan-Meier analyses of overall survival and disease-free survival, respectively. ^aP < 0.05, ^dP < 0.0001. AUC: Area under the curve; NCRT: Neoadjuvant chemoradiotherapy.

Correlation between Linc00909 expression and clinicopathological characteristics

The patients were divided into high- and low-expression groups based on the median expression level of Linc00909 (7.98) in 196 colorectal cancer tissues. χ² tests and t-tests revealed a significant positive correlation between Linc00909 expression and pM stage (P < 0.001), but no significant associations were detected with sex, age, American Society of Anesthesiologists score, or tumor location (Table 1).

Table 1 Baseline characteristics of the 196 patients with colorectal cancer.

Variable	Linc00909 expression		P value
	High (n = 98)	Low (n = 98)
Sex			0.770
Male	58 (59.18)	60 (61.22)
Female	40 (40.82)	38 (38.78)
Age, years	61.08 ± 12.08	60.60 ± 11.60	0.777
American Society of Anesthesiologists score			0.793
1	16 (16.33)	19 (19.39)
2	71 (72.45)	70 (71.43)
3	11 (11.22)	9 (9.18)
Tumor location			0.997
Ascending colon	15 (15.31)	15 (15.31)
Transverse colon	10 (10.20)	9 (9.18)
Descending colon	6 (6.12)	7 (7.14)
Sigmoid colon	21 (21.43)	20 (20.41)
Rectum	46 (46.94)	47 (47.96)
Gross type			0.125
Expanding	24 (24.49)	36 (36.73)
Ulcerating	73 (74.49)	61 (62.25)
Infiltrating	1 (1.02)	1 (1.02)
Histopathology			0.461
Adenocarcinoma	78 (79.59)	82 (83.67)
Mucinous adenocarcinoma	20 (20.41)	16 (16.33)
Tumor differentiation			0.902
Well differentiated	4 (4.08)	4 (4.08)
Moderately differentiated	70 (71.43)	74 (75.51)
Poorly differentiated	19 (19.39)	15 (15.31)
Undifferentiated	5 (5.10)	5 (5.10)
Postoperative hospital stay, days	9.11 ± 7.36	8.02 ± 6.80	0.282
Number of lymph nodes retrieved	22.57 ± 12.96	24.04 ± 15.35	0.470
Number of metastatic lymph nodes	1.70 ± 3.46	1.37 ± 2.98	0.466
Tumor size, cm	4.47 ± 2.22	4.20 ± 2.19	0.402
Pathological T stage			0.367
T1	7 (7.14)	5 (5.10)
T2	17 (17.35)	19 (19.39)
T3	60 (61.22)	67 (68.37)
T4	14 (14.29)	7 (7.14)
Pathological N stage			0.214
N0	53 (54.08)	65 (66.32)
N1	28 (28.57)	20 (20.41)
N2	17 (17.35)	13 (13.27)
Pathological M stage			< 0.001
M0	71 (72.45)	93 (94.90)
M1	27 (27.55)	5 (5.10)
Neural invasion			0.610
Yes	21 (21.43)	24 (24.49)
No	77 (78.57)	71 (75.51)
Vascular invasion			0.809
Yes	9 (9.18)	10 (10.20)
No	89 (90.82)	88 (89.80)

Values are presented as n (%) or mean ± SD. P values were calculated using the χ² test for categorical variables and Student’s t-test for continuous variables. Statistical significance was set at P < 0.05.

Cox regression analysis

To further determine the factors influencing patient prognosis, we conducted univariate and multivariate Cox regression analyses. Univariate Cox regression revealed that gross tumor classification (P = 0.032), pT stage (P = 0.005), pN stage (P = 0.002), nerve invasion (P = 0.003), and Linc00909 expression level (P = 0.007) were significantly associated with OS. Furthermore, gross tumor classification (P = 0.036), pT stage (P = 0.001), pN stage (P < 0.001), number of lymph node metastases (P < 0.001), nerve invasion (P = 0.049), vascular tumor thrombus (P = 0.047), and Linc00909 expression level (P < 0.001) were significantly associated with DFS. Multivariate analysis confirmed that nerve invasion (hazard ratio = 3.109, 95% confidence interval: 1.245-7.764, P = 0.015) and Linc00909 expression (hazard ratio = 5.067, 95% confidence interval: 1.629-15.765, P = 0.005) were independently associated with OS. High Linc00909 expression was also correlated with poorer DFS (hazard ratio = 5.335, 95% confidence interval: 2.467-11.536; P < 0.001) (Tables 2 and 3).

Table 2 Univariate and multivariate Cox regression analyses of overall survival in the 196 patients with colorectal cancer.

Variable	Univariate analysis			Multivariate analysis
	HR	95%CI	P value	HR	95%CI	P value
Sex, male/female	1.176	0.487-2.838	0.719
Age	0.998	0.962-1.036	0.918
American Society of Anesthesiologists	1.092	0.471-2.530	0.837
Postoperative hospital stay	1.011	0.959-1.065	0.695
Tumor size	1.070	0.889-1.287	0.474
Histopathology	0.459	0.106-1.980	0.296
Tumor location			0.686
Ascending colon	Reference	Reference
Transverse colon	1.802	0.253-12.806	0.556
Descending colon	1.168	0.106-12.901	0.899
Sigmoid colon	1.667	0.305-9.105	0.555
Rectum	2.028	0.449-9.152	0.358
Tumor differentiation	0.358	0.083-1.544	0.169
Gross type	0.112	0.015-0.827	0.032	0.188	0.025-1.427	0.106
Pathological T stage	2.928	1.372-6.246	0.005	1.371	0.616-3.053	0.439
Pathological N stage	2.293	1.369-3.841	0.002	1.755	0.987-3.120	0.055
Lymph nodes retrieved	0.982	0.946-1.018	0.326
Metastatic lymph nodes	1.083	0.996-1.178	0.062
Neural invasion	3.775	1.570-9.076	0.003	3.109	1.245-7.764	0.015
Vascular invasion	0.524	0.070-3.925	0.529
Linc00909 expression	4.518	1.510-13.522	0.007	5.067	1.629-15.765	0.005

Univariate analysis was performed for each variable; variables with P < 0.05 in the univariate analysis were entered into the multivariate Cox proportional hazards model (using forward selection). Bold P values indicate statistical significance (P < 0.05). The reference group for Linc00909 expression was low expression. CI: Confidence interval; HR: Hazard ratio.

Table 3 Univariate and multivariate Cox regression analyses of disease-free survival in the 196 patients with colorectal cancer.

Variable	Univariate analysis			Multivariate analysis
	HR	95%CI	P value	HR	95%CI	P value
Sex, male/female	0.889	0.496-1.596	0.694
Age	1.004	0.980-1.029	0.754
American Society of Anesthesiologists score	0.971	0.563-1.674	0.916
Postoperative hospital stay	0.984	0.938-1.033	0.516
Tumor size	0.949	0.829-1.087	0.451
Histopathology	0.855	0.400-1.827	0.687
Tumor location			0.686
Ascending colon	Reference	Reference
Transverse colon	0.690	0.178-2.667	0.590
Descending colon	0.656	0.136-3.160	0.599
Sigmoid colon	0.952	0.354-2.555	0.922
Rectum	1.312	0.571-3.013	0.522
Tumor differentiation	0.992	0.506-1.945	0.981
Gross type	0.466	0.228-0.952	0.036	0.681	0.315-1.476	0.331
Pathological T stage	2.202	1.374-3.528	0.001	1.565	0.945-2.593	0.082
Pathological N stage	1.941	1.389-2.712	< 0.001	1.127	0.688-1.846	0.636
Lymph nodes retrieved	0.987	0.957-1.004	0.110
Metastatic lymph nodes	1.115	1.059-1.174	< 0.001	1.069	0.986-1.159	0.104
Neural invasion	1.827	1.002-3.333	0.049	1.365	0.703-2.650	0.357
Vascular invasion	2.164	1.009-4.640	0.047	1.450	0.613-3.429	0.397
Linc00909 expression	5.643	2.638-12.070	< 0.001	5.335	2.467-11.536	< 0.001

Univariate analysis was performed for each variable; variables with P < 0.05 in the univariate analysis were entered into the multivariate Cox proportional hazards model (using forward selection). Bold P values indicate statistical significance (P < 0.05). High Linc00909 expression was used as the reference category. CI: Confidence interval; HR: Hazard ratio.

Role of Linc00909 in mediating rectal cancer cell proliferation and apoptosis in vitro

To investigate the role of Linc00909 in cell proliferation, we stably overexpressed Linc00909 in the SW620 and DLD1 cell lines. Successful overexpression was confirmed at the RNA level in both lines (Figure 2A). CCK-8 assays revealed that Linc00909 overexpression promoted cell proliferation (Figure 2B). This effect was further corroborated by clonogenic assays (Figure 2C and D). We next examined the effect of Linc00909 on apoptosis in colorectal cancer cells. Flow cytometry revealed that, compared with the control conditions, Linc00909 overexpression significantly reduced the percentage of apoptotic cells in both the SW620 and DLD1 cell lines (Figure 2E and F).

Figure 2 Role of Linc00909 in regulating rectal cancer cell proliferation and apoptosis in vitro. A: Linc00909 overexpression efficiency was measured by real-time quantitative polymerase chain reaction; B: Cell proliferation was assessed using a Cell Counting Kit-8 assay; C and D: Colony formation assays were performed to assess the proliferative capacity of the cells; E and F: Apoptosis was analyzed by flow cytometry. ^aP < 0.05, ^bP < 0.01, ^cP < 0.001. CON: Control; OE: Overexpression; VEC: Vector control.

Effects of Linc00909 on the cell cycle distribution and sphere-forming capacity of rectal cancer cells in vitro

To further investigate the impact of Linc00909 on the cell cycle, flow cytometry was used to analyze the cell cycle distribution of the Linc00909-overexpressing colorectal cancer cell lines SW620 and DLD1. In SW620 cells, Linc00909 overexpression resulted in a marked reduction in the number of G1-phase cells and a concomitant increase in the S- and G2/M-phase populations compared with those in the vector and control groups. A consistent phenotype was observed in DLD1 cells (Figure 3A and B). To further investigate the role of Linc00909 in colorectal cancer cells, we examined its effect on tumor sphere formation using a sphere-forming assay following Linc00909 overexpression. As shown in Figure 3C, Linc00909 overexpression in SW620 cells significantly increased the diameter of the tumor spheres compared with those in the vector control and blank control groups. A similar trend was observed in DLD1 cells (Figure 3D).

Figure 3 Linc00909 regulates cell cycle progression and sphere formation in rectal cancer cells in vitro. A and B: Cell cycle distribution across all cell lines was analyzed by flow cytometry; C and D: Sphere formation assays were performed to evaluate the stem cell-like properties of the colorectal cancer cell lines. ^aP < 0.05, ^bP < 0.01, ^cP < 0.001. 5-FU: 5-Fluorouracil; CON: Control; OE: Overexpression; RT: Radiotherapy; VEC: Vector control.

Effect of Linc00909 on 5-FU and radiotherapy sensitivity in rectal cancer cells

To evaluate the role of Linc00909 in regulating 5-FU and radiotherapy sensitivity in rectal cancer cells, cell viability was assessed using CCK-8 assays following treatment. Based on the results, the half-maximal inhibitory concentration (IC₅₀) values for 5-FU and radiotherapy were calculated in SW620 cells overexpressing Linc00909, empty vector controls, and untreated controls. The IC₅₀ values for 5-FU were 23.75 μmol/L, 2.12 μmol/L, and 4.28 μmol/L, respectively, whereas those for radiotherapy were 39.05 Gy, 6.86 Gy, and 10.85 Gy, respectively (Figure 4A and B). In DLD1 cells, the IC₅₀ value for 5-FU was 71.48 μmol/L in the Linc00909-overexpressing group, whereas it was 12.35 μmol/L in the vector control group and 15.78 μmol/L in the untreated control group. The corresponding IC₅₀ values for radiotherapy were 32.92 Gy, 13.30 Gy, and 12.84 Gy, respectively (Figure 4A and B). These results revealed that, compared with control cells, Linc00909-overexpressing cells exhibited significantly reduced sensitivity to 5-FU- and radiotherapy-induced proliferation inhibition. Additionally, the apoptosis rates of Linc00909-overexpressing, vector control, and untreated control cells following 5-FU or radiotherapy treatment were analyzed using flow cytometry. As shown in Figure 4C and D, Linc00909-overexpressing SW620 and DLD1 cells exhibited significantly lower apoptosis rates following 5-FU or radiotherapy than did the vector control cells.

Figure 4 Linc00909 modulates the sensitivity of rectal cancer cells to 5-fluorouracil and radiotherapy. A: Sensitivity to 5-fluorouracil was assessed using a Cell Counting Kit-8 assay in all cell lines; B: Sensitivity to radiation was assessed using a Cell Counting Kit-8 assay in all cell lines; C and D: Radiosensitivity across all cell lines was assessed using a Cell Counting Kit-8 assay. Apoptosis in SW620 cells following treatment with 5-fluorouracil and radiotherapy was analyzed by flow cytometry (C). Apoptosis in DLD1 cells following treatment with 5-fluorouracil and radiotherapy was analyzed by flow cytometry (D). ^aP < 0.05, ^bP < 0.01, ^cP < 0.001, ^dP < 0.0001. CON: Control; OE: Overexpression; VEC: Vector control.

Effect of Linc00909 overexpression on subcutaneous xenograft tumors derived from colorectal cancer cells in nude mice

To investigate the in vivo effect of Linc00909 overexpression on colorectal cancer cell proliferation, a subcutaneous xenograft model was established in nude mice. Linc00909 overexpression significantly increased tumor volume in both SW620 and DLD1 cell-derived xenografts compared with that in the vector control groups (Figure 5A and B).

Figure 5 Linc00909 overexpression promotes the growth of colorectal cancer cell-derived subcutaneous xenografts in nude mice. A and B: Linc00909 overexpression increased subcutaneous xenograft tumor volume in nude mice. ^aP < 0.05. OE: Overexpression; VEC: Vector control.

Prediction of miRNAs that interact with Linc00909 and their downstream target genes via bioinformatic analysis

To identify downstream molecules that interact with Linc00909, we performed cluster analysis of previously obtained microarray data from chemoradiation-sensitive and chemoradiation-resistant groups, selecting the most differentially expressed genes. Differentially expressed lncRNAs were subsequently subjected to Gene Ontology annotation and Kyoto Encyclopedia of Genes and Genomes pathway enrichment analysis. As shown in Figure 6A and B, Linc00909 was significantly associated with adenosine triphosphate-binding cassette transporters and the MAPK, WNT, and JAK-STAT signaling pathways. Bioinformatic analysis using TargetScan, miRcode, miRDB, and miRTarBase predicted potential miRNA binding sites on Linc00909 and its downstream target genes (Figure 6C). The stemness pathway marker SOX2 was shown to interact with both Linc00909 and miR-625-5p. In colorectal cancer, SOX2 dysregulation is associated with chemoresistance and radioresistance. We found that Linc00909 overexpression enhanced stemness features and modulated the sensitivity of tumor cells to chemoradiation. We then quantified miR-625-5p and SOX2 expression by RT-qPCR and SOX2 protein expression by western blotting in colorectal cancer cells following Linc00909 overexpression. Linc00909 overexpression significantly downregulated miR-625-5p and upregulated SOX2 expression at both the RNA and protein levels in SW620 and DLD1 cells, as shown by RT-qPCR and western blotting (Figure 6D-I). These results suggest that Linc00909 regulates SOX2 expression by sequestering miR-625-5p.

Figure 6 Bioinformatic prediction of microRNAs that interact with Linc00909 and their downstream target genes. A and B: Gene Ontology and Kyoto Encyclopedia of Genes and Genomes enrichment analyses were performed based on the significantly differentially expressed genes identified between the chemosensitive and chemoresistant groups; C: Bioinformatic analysis was used to predict microRNAs that may bind to Linc00909 and their downstream target genes; D and E: The expression levels of miR-625-5p and SOX2 were quantified by real-time quantitative polymerase chain reaction; F-I: SOX2 protein expression was analyzed by western blotting. ^aP < 0.05, ^cP < 0.001. CON: Control; OE: Overexpression; VEC: Vector control.

DISCUSSION

Colorectal cancer ranks as the third most common malignancy globally, with its incidence and mortality steadily increasing in China over the past decade, representing a major public health challenge. Although lncRNAs were initially considered nonfunctional transcriptional byproducts, accumulating evidence has demonstrated their critical roles in regulating gene expression at the transcriptional, pretranscriptional, and posttranscriptional levels, thereby modulating tumor initiation, progression, metastasis, and recurrence in colorectal cancer and other cancers[26-28]. For example, Fan et al[29] reported that the lncRNA MALAT1 acts as an oncogene in colorectal cancer, where its upregulation enhances cell migration and confers resistance to oxaliplatin. Similarly, PVT1 is highly expressed in non-small cell lung cancer and is positively correlated with tumor size, lymph node metastasis, histological grade, and tumor-node-metastasis stage, indicating poor prognosis[30-32].

Emerging studies further suggest that lncRNAs influence chemoradiation sensitivity in colorectal cancer by functioning as ceRNAs or molecular sponges for miRNAs, thereby interacting with proteins and mRNAs to form complex regulatory networks. Wang et al[22] demonstrated that HNF1A-AS1 sequesters miR-92a-3p to upregulate MAP2K4, activating the MAPK/JNK pathway and modulating radiosensitivity in non-small cell lung cancer.

Although lncRNAs are increasingly implicated in therapy resistance, their mechanistic roles in the response of rectal cancer to chemoradiation remain incompletely elucidated. In this study, we found that Linc00909 expression was significantly elevated in neoadjuvant chemoradiation-resistant tissues compared with sensitive tissues, as well as in tumor tissues relative to adjacent normal tissues. High Linc00909 expression was associated with poorer patient prognosis. Collectively, these findings suggest that Linc00909 may act as an oncogenic driver in colorectal cancer and serve as a potential biomarker for predicting patient prognosis and therapeutic response.

To further investigate the role of Linc00909 in tumorigenesis and progression, we noted that it is highly expressed in human glioma and functions as an oncogene. Linc00909 acts as a molecular sponge for miR-194, competitively increasing the expression of the downstream target MUC1-C and promoting glioma development[24]. In colorectal cancer, Xu et al[25] identified high serum levels of Linc00909 as a potential diagnostic biomarker; however, the functional role of Linc00909 in colorectal cancer pathogenesis, particularly its impact on chemoradiotherapy sensitivity, remains unexplored. In this study, we established stable Linc00909-overexpressing colorectal cancer cell lines (SW620 and DLD1) using lentiviral vectors and performed a series of in vitro and in vivo experiments. Linc00909 overexpression significantly increased proliferative capacity, colony formation, apoptosis resistance, and sphere-forming ability; promoted G1- to S/G2-phase cell cycle transition; and accelerated tumor growth in nude mice, which is consistent with its oncogenic role in colorectal cancer pathogenesis. Furthermore, Linc00909 overexpression markedly increased resistance to 5-FU and radiotherapy, supporting its involvement in regulating chemoradiotherapy sensitivity. However, the absence of Linc00909-knockdown models limits the comprehensiveness and mechanistic depth of our functional analysis.

Our findings indicate that Linc00909 is closely associated with malignant phenotypes, stem-like properties, and chemoradiotherapy sensitivity. However, the precise mechanistic pathways and downstream targets through which it performs these functions remain unclear. Bioinformatic analyses revealed that miR-625-5p may bind to Linc00909 and that SOX2 is a potential target of miR-625-5p. Linc00909 overexpression significantly downregulated miR-625-5p and upregulated SOX2 expression in SW620 and DLD1 cells, as confirmed by RT-qPCR and western blotting. These results suggest that Linc00909 may function as a ceRNA by sequestering miR-625-5p, thereby regulating SOX2 expression. SOX2 is a transcription factor featuring a high-mobility group domain and sequence-specific DNA-binding activity[33,34]. Its expression is elevated in multiple cancers and is correlated with poor prognosis[35-38]. SOX2 plays a central role in maintaining tumor-initiating cells, governing self-renewal, and sustaining tumorigenic potential across diverse cancer types[39,40]. In colorectal cancer, SOX2 is highly expressed in tumor tissues and cell lines, where its dysregulation is associated with aggressive phenotypes, including metastasis, chemoresistance, and radioresistance, and serves as an indicator of unfavorable patient outcomes[41-43]. We found that Linc00909 overexpression upregulated SOX2 expression and enhanced stem-like properties and chemoradiotherapy resistance. However, our study only indirectly suggested that Linc00909 modulates these phenotypes through SOX2 regulation, and further functional validation of the biological roles and mechanistic pathways of SOX2 in colorectal cancer remains necessary.

CONCLUSION

Linc00909 modulates colorectal cancer cell proliferation, colony formation, apoptosis, cell cycle progression, sphere-forming ability, and sensitivity to 5-FU and radiotherapy. Its overexpression also promotes tumor growth in nude mice. Mechanistically, Linc00909 acts by sponging miR-625-5p to upregulate SOX2 expression, thereby enhancing cancer stem-like properties and contributing to chemoradiotherapy resistance. Therefore, Linc00909 represents a potential therapeutic target in colorectal cancer.